You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When running the automatic database setup using docker-compose run db-setup as in the README.md the download is terminated prematurely with some text output stating for ecample:
Cannot write to '/tmp/12645459548716115717/highquality_clust30_ca' (No space left on device).
This is caused because the download script stores files in the /tmp directory of the container and docker's default maximum tmpfs size being 50 % of the RAM available on the host machine (which in my case was 32 G).
As stated in the docker documentation when using docker run you can control the upper limit of the tmpfs mount using the --tmpfs-size option however in docker-compose this can be set like:
tmpfs:
- /tmp:exec,size=200G
This however did not work for me, the container crashed every time I tried. Instead the workaround I used was to simply change the directory the downloads were sent to from /tmp to /downloads, lines 11 and 15 of setup_db.sh
setup_db.sh Line 11
When running the automatic database setup using
docker-compose run db-setup
as in the README.md the download is terminated prematurely with some text output stating for ecample:This is caused because the download script stores files in the /tmp directory of the container and docker's default maximum tmpfs size being 50 % of the RAM available on the host machine (which in my case was 32 G).
As stated in the docker documentation when using docker run you can control the upper limit of the tmpfs mount using the
--tmpfs-size
option however in docker-compose this can be set like:This however did not work for me, the container crashed every time I tried. Instead the workaround I used was to simply change the directory the downloads were sent to from /tmp to /downloads, lines 11 and 15 of setup_db.sh
setup_db.sh Line 11
setup_db.sh Line 15
Maybe someone else has a more elegant solution?
The text was updated successfully, but these errors were encountered: